Magnetic resonance imaging of Huntington's disease: preparing for clinical trials

Abstract

The known genetic mutation causing Huntington's disease (HD) makes this disease an important model to study links between gene and brain function. An autosomal dominant family history and the availability of a sensitive and specific genetic test allow pre-clinical diagnosis many years before the onset of any typical clinical signs. This review summarizes recent magnetic resonance imaging (MRI)-based findings in HD with a focus on the requirements if imaging is to be used in treatment trials. Despite its monogenetic cause, HD presents with a range of clinical manifestations, not explained by variation in the number of CAG repeats in the affected population. Neuroimaging studies have revealed a complex pattern of structural and functional changes affecting widespread cortical and subcortical regions far beyond the confines of the striatal degeneration that characterizes this disorder. Besides striatal dysfunction, functional imaging studies have reported a variable pattern of increased and decreased activation in cortical regions in both pre-clinical and clinically manifest HD-gene mutation carriers. Beyond regional brain activation changes, evidence from functional and diffusion-weighted MRI further suggests disrupted connectivity between corticocortical and corticostriatal areas. However, substantial inconsistencies with respect to structural and functional changes have been reported in a number of studies. Possible explanations include methodological factors and differences in study samples. There may also be biological explanations but these are poorly characterized and understood at present. Additional insights into this phenotypic variability derived from study of mouse models are presented to explore this phenomenon.

Fig. 1

The figure illustrates the impact of different smoothing kernels (ranging from 10 to 30 mm) on the apparent distribution and severity of cortical thinning.

Fig. 2

Example displays areas with decreased GM (P<0.001) as detected by VBM comparing early HD subjects to controls (see Kassubek et al., 2005; for details on group characteristics and methodology).

Fig. 3

Positive TCN identified by means of ICA in healthy controls (green) and pre-HD subjects (blue) during the delay period of a verbal working memory task (Wolf and Walter, 2005; Wolf et al., 2008a). Top right: Within a predominantly frontostriatal TCN, areas with decreased connectivity in pre-HD individuals versus healthy comparison subjects comprised the left ventro- and dorsolateral prefrontal cortex, the left parietal lobule, the left insula, the bilateral putamen and the right caudate. Bottom right: Within a predominantly frontoparietal TCN, areas with decreased connectivity in pre-HD individuals versus healthy comparison subjects comprised the left dorsolateral prefrontal cortex and the left superior parietal cortex (Wolf et al., 2008b). The 2nd level spatial maps are rendered on the anatomical templates implemented in MRIcron [http://www.sph.sc.edu/comd/rorden/mricron/] (left) and SPM5 (right).